This week has once again seen the assessment of energy expenditure using food diaries heavily criticised during the peer review process. I guess everyone who has had the misfortune and suffered the pain of collecting energy intake data and then analysing it using dietary analyses packages, understands the pitfalls and inaccuracies of energy intake; but given the importance of assessing energy intake it is still something most researchers (and practitioners) strive to achieve. However, has the time come to accept that the inaccuracies are so great that ultimately what we are often doing is wasting our time, but perhaps more importantly, wasting our athletes time – and given we often have very limited time with athletes, could this time be spent more wisely?
Accurate assessment of Energy Intake is extremely important – but is it possible to do it?
The Typical Process
When I first started to consult in sport nutrition I spent a lot of time designing a food diary, and this formed the basis of my consultations. I still remember consultations where I excitedly gave my well-designed diary to athletes and then saw the disappointment in their faces. After all, who really wants to write down everything they are going to eat over the next 3-7 days – or even worse weigh it all. But I persisted and asked the athletes to complete the diary for 3-7 days and write down everything they eat or drink.
My next observation was that typically I would rarely get a food diary back from the athletes and when I did it was very clear that this was not the most accurate account of their intake.
Once we get the diary the next job is to analyse it. This can take several hours of inputting food items into various software packages. However, sometimes the food item you are looking for is not there. And what do we do with “homemade lasagne” or Chinese take away Chow Mein. If we get food descriptors like a large chicken breast and large potato and 1 medium carrot this is reasonable easy to analyse but let’s be honest how many of our athletes eat like this. And the ones that do probably are not the ones that need us to analyse their food diary.
Eventually, we get a nice report from our software of a day by day breakdown of calories (to the nearest Kcal, protein, carbs, fats to the nearest gram) and even micronutrients to the nearest milligram. We then proudly take this report back to our athlete – but what have we given them? Despite having extremely precise numbers, do they actually mean anything?
A brilliant researcher on energy intake from LJMU (Allan Hacket) once said to me that assessing energy intake is the most difficult of all physiological measurements. At first I thought he had gone mad, surely measuring what someone eats must be easier than measuring muscle glycogen for example. However, after some thought, and many attempts to do this I now 100% agree. Think about it, to measure muscle glycogen we follow a process and if we do this correctly we get an accurate number. But with energy intake you could do everything correct yourself but the data could still be nonsense. We are completely reliant on our athlete’s ability to record data well and perhaps more importantly being 100% honest.
The literature tells us that energy intake is subject to both deliberate and accidental under reporting with numbers such as 30% often cited. Experience from our laboratory would confirm that these numbers are correct, if not slightly underestimating the extent of the problem. We have used doubly labelled water to assess energy expenditure in both jockeys and rugby players whilst at the same time measuring energy intake using food diaries and diet recalls. The results can be seen below in the 2 slide images.
Above – Jockey’s energy expenditure and intake. Blue bar = resting metabolic rate, green bar = total energy expenditure and yellow bar = self reported energy intake.
Above- Energy Intake and expenditure from professional rugby players. Note that with the exception of one every player under reported energy intake.
So why are food diaries so inaccurate? I guess there are many reasons but perhaps the most obvious ones are:
- They simply do not want to fill them in so do it rushed and without care. Or forget to fill it in and then try and guess at the end of the collection period.
- Given it is a written document often players do not want to write down poor food choices for fear of punishment from coaching staff.
- Establishing portion sizes is extremely hard to do, players often do not want to weigh food and using descriptors like big, small, handful etc can be inaccurate.
- We don’t ask the athlete to complete them for long enough. Often, we use 3-5 days for convenience but we know that for things like vitamin C this may need up to 36 days to get accurate data.
- Athletes may change what they eat. So, you may have really accurate account of what they ate for those 3 days but this is not representative of what they usually eat.
- Even if we use pictures to help record data, is an athlete going to take a picture of a donner kebab, chips and lager to send back to the club if they are being told they must lose weight? Absolute honesty is essential from athletes if we are even going to attempt to assess intake. And can a picture always tell you want you want. Think of the calorific difference between a thin spread of nut butter (65 Kcal) and a thick spread (perhaps 300Kcal). Can we really work this out from a picture?
Above – Number of days needed to collect a food diary to get accurate information for various measures.
Having said all this, we have recently collected some good food diaries on young professional football players. James Morton was the lead author in this paper and puts the excellent data down to the close bond the PhD student had with the players. Similarly, there was a brilliant study on Irish Jockeys from Giles Warrington’s group where their energy intake data looks extremely believable and matched our expenditure data collected using DLW. The reason for this great data; the jockeys were all wearing a portable camera! This is perhaps the best way to collect food diary data.
- The first thing I suggest is that we need to re-read the data on energy intake on athletes, but with a huge pinch of salt. I’ve seen papers suggesting that athletes attempting to lose weight are not eating enough based on intake data where in reality we all know that they are not losing weight as they are eating too much. Similarly, we see lots of studies suggesting that athletes fail to meet their carbohydrate needs where in reality there is a good chance they simply under reported. Some journals are even calling now for all food intake data to be scrapped from studies given it has led to such errors in the literature.
- If we are going to use food diaries in research we need to think carefully about how to get reliable data. Building a great rapport with the athletes is essential, but we can also look at taking pictures and wearing cameras. This could still change their behaviour hence the good rapport is the starting point for all of this.
- In applied practice think carefully if there is any point asking an athlete to complete a food diary. Sometimes a quick diet recall can be much better. This way you can give immediate feedback, the player does not feel like they are recording in writing bad foods and you can tease out more information. This also helps to build the rapport and you can get some very quick wins in these meetings. It may take a week to get some information back where with a recall the athlete can leave the consultation with feedback and a plan of action.
- I tend to use a food diary only to assess patterns or in very enthusiastic athletes. I do not these days even use diet analysis software at the end. This has saved me literally hours of time and allowed me to spend this precious time much more effectively speaking with athletes and putting interventions into action. I am more interested in the timing, type and rough total of food intake. Are they eating protein regularly, are there vegetables in the diet, do the carbohydrate sources look to be good quality etc. The actual numbers do not over worry me as even if they are very carefully calculated do we even have accurate assessments of the player’s energy expenditure to match this to?
- One use for them could be to assess honesty and knowledge in your athletes. If you have an overweight athlete where the food diary comes back amazing, you have a good guess they know what to eat but are just not doing it. This then requires behaviour change (see blog 4). However, if there are pie and chips (or a pie barm) all over the diary, then there is a good chance you need to focus your time on education.
- Technology may be able to help us a little. Apps like ‘My Fitness Pal’ with bar scanning technology can simplify the process plus athletes tend to take their phones everywhere. There is also technology on the horizon that may help and I covered this recently in a talk at ECSS which you can watch here if you have 30 mins to spare.
- And finally let’s get back to first principles. We may think the food diary is brilliant and shows us that our player is on low enough calories intake to lose weight. But if they are not losing weight, it does not matter what the food diary is telling you, they are eating too much. Like all things in sport science, do not let your numbers distract you from the end prize. We need to be able to see that proverbial wood from the trees.
OK, promised a shorter one this week. In summary, I do not quite think the food dairy has been served its last supper yet, but we need to treat them with a great deal of caution and perhaps change how we are using them. As a sport nutritionist, our interactions (and indeed our time) with the athletes is very precious so we must always make sure we are using it wisely – I am not convinced that asking most athletes to complete a diary over 7 days and then us punching numbers into a computer is a wise use of our time. Although, this is just my opinion and experience, others may find them much more effective than me. If you do please leave a comment below and share ideas.
Until the next Close Encounter of the Nutritional Kind